More About Bunting and Beating the Shift

Apparently what I’m really interested in these days is bunting. Which is a little odd, but which is also a welcome break from thinking about pitch-framing all the time. And though bunting is ordinarily associated with the dreaded and unproductive sacrifice, a well-placed bunt in fair territory can make for a hell of a weapon, in particular when it’s put down by a left-handed hitter against a defensive overshift. Everybody’s talking about shifting these days, and while it’s hard for a hitter to make dramatic changes to his swing, oftentimes there’s a single there for the taking, and it would require no swing at all. So bunting and the shift are worth analyzing in more detail.

Yesterday, I was excited to generate some bunting statistics I’d never seen anywhere before. Obviously, I’d seen bunt batting averages and whatnot, but I’d never seen attempts, including foul bunts and missed bunts. I was pleased to have that data, but the data also didn’t say enough. You can’t just characterize a fair bunt as a successful bunt. And within all bunt attempts are attempts at different sorts of bunts. A sacrifice bunt is different from a bunt for a hit, and a bunt for a hit against the shift is different from a bunt for a hit against a regular alignment. I made it a goal to dig a little deeper, because, what is it to bunt against the shift, really?

The first thing I did was pull up this article from the Hardball Times, talking about hitters shifted most often. That mostly confirmed what I would’ve expected. Then I started generating numbers. I looked for bunt attempts by lefties with nobody on base. Then I narrowed the numbers down, based on who I figured was bunting against the shift, and who was just bunting for an ordinary attempted single. The former group was mostly plodders and sluggers — the latter group was mostly speedsters. This required some judgment calls, because I didn’t confirm every single attempt on video, but while the numbers might not be perfect, they ought to be close. I generated numbers for 2013, and then I did the same for 2012, following the same guidelines and assumptions.

It won’t surprise you to learn there haven’t been all that many bunt attempts against the shift. This is the very reason certain people have been complaining. I came up with just over 200 attempts over two years. Of those attempts, 38% were bunted fair, and 25% of the bunts resulted in the batter reaching base, either on a hit or an error. In other words, one of four attempted bunts put the batter on, but two of three bunts in play worked out, which isn’t a shock. Thing about the overshift is there’s not really anyone in the vicinity to do the fielding.

You might think these rates look low. Or you might think these rates look high! We don’t really have much of a frame of reference. What we can guess is that these are the rates by and large posted by a group of hitters who don’t have much experience bunting. So with more practice, we’d expect some improvement. Here’s Athletics batter Brandon Moss, who frequently faces a shift:

“A lot of guys might look at it as, ‘If I bunt, I’m not giving myself a chance to drive the ball.’ You’re just giving yourself an opportunity to get on base for your team. It was something I wanted to do last year, but I had no idea what I was doing. I think they want me to do it, too.”

Moss acknowledges that bunting was weird. He’s actively working this spring training to make bunting less weird. What one would expect is that Moss would become better at bunting successfully. And then that’s just something to have in his back pocket.

This is what a regular successful bunt against the shift looks like:

MorrisonBunt.gif.opt

And this is how easy it is to reach base if you can put the ball where you want:

CongerBunt.gif.opt

What might be the upside for shifted hitters learning to bunt? The undisputed king of bunting to beat the shift is Carlos Pena. Since 2008, with nobody on, Pena has attempted a bunt 65 times. He’s bunted the ball fair 33 times, and he’s reached base 23 times. Even with practice, Pena’s right around a 50/50 fair/not-fair rate, but he’s turned better than a third of his attempts into bases. We find Jay Bruce at 30 attempts, with nine fair and five successful. Brian McCann also has 30 attempts, with eight fair and six successful. What’s clear is that bunting against the shift isn’t automatic. What’s also clear is that it’s worked often, and that these hitters could do better if they just practiced their bunting more. I suppose that’s just a guess, but I feel good about it.

What does bunting in front of a shifted defense do to the defense? After all, there are two thoughts — not only can a bunt mean a base, but it can also mean an adjusted shift that might make it more productive to swing away. It might even make the shift go away altogether. Here’s Chris Davis showing bunt against the Angels last May:

davisbunt0

In the next game, Davis tried a bunt. Here’s the infield:

davisbunt1

Overshift. Davis then tried to bunt again. Here’s the infield:

davisbunt2

The third baseman had moved in a little, but the shift was still on. Here’s the infield after two bunt attempts, along with the shown bunt the previous evening:

davis2strikes

Still shifted. The third baseman had cheated over a little, but only by a matter of steps. And as it happened, Davis hit a grounder right to him. The Angels acknowledged the bunt possibility, but they did little to counter it.

How about Adrian Gonzalez, last July? Here’s Gonzalez attempting a bunt:

gonzalezbunt1

Extreme shift. Gonzalez tried to bunt again, immediately:

gonzalezbunt2

Same shift. Here’s the infield after the two bunt attempts:

gonzalezbunt3

No changes. In fairness, Gonzalez didn’t do much of anything to force a change — his bunts, after all, were unsuccessful. You’d think a manager would only change his shift if he felt like there was an actual bunt threat. Why don’t we check in on something, then? Carlos Pena, as noted, is the king of bunting against the shift. Pena played a long time with Tampa Bay. Here’s Pena facing Tampa Bay last season, right before Pena was dropped:

penaraysshift

Extreme shift. The Rays, of course, hadn’t forgotten about Pena’s bunting ability. But that picture was taken with two strikes, and with two strikes, the Rays changed things up, swapping Yunel Escobar and Evan Longoria. Here’s how the Rays played Pena before two-strike counts:

penaraysshift2

penaraysshift3

Still three shifted infielders, but one in close, protecting against the possible bunt attempt. Pena’s history didn’t cause the Rays to abandon the shift — it just caused them to alter it a little, moving one guy closer in and closer to third. That’s also the one guy least likely to have a batted ball hit on the ground in his direction anyway. The message being, bunts don’t make the shift disappear. At least, there’s no evidence of that yet. They might just make the shifts look a little bit different, but most of the action’s still going to go between first and second base. There might be more area opened up around where the shortstop would usually be, but lefties don’t hit a ton of grounders over there, and there is still a defender on that side no matter what.

The math is still probably on the side of more bunt attempts. It’s definitely on the side of hitters practicing bunting more often so that they can be more successful when it counts. There should at least be the perception of the threat of a bunt. Over the last two years, shifted batters haven’t been great bunters, but when they’ve bunted the ball fair they’ve reached base twice as often as they haven’t. The numbers show that bunting against the shift is by no means an easy thing, not that any aspect of hitting is easy. And based on the Carlos Pena/Rays example, it doesn’t look like bunting is the solution to eliminating the overshift. It might just slightly alter the overshift, if bunts are dropped down with greater success and greater frequency. Shifting, presumably, is here to stay. The only question is exactly what the shifts are going to look like.



Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


Sort by:   newest | oldest | most voted
Eric
Guest
Eric
2 years 2 months ago

I don’t care about bunting, but I would love it if fangraphs would answer its email bag once in awhile. I sent two emails and NEVER heard back.

Canucklesndwch
Guest
Canucklesndwch
2 years 2 months ago

what do you care about Eric? would you like a feature on what john Rocker is up to these days? Or a thorough examination of what types of grass is being used at MLB stadiums? Now’s your chance – tell the world how smart you are!

Spit Ball
Guest
Spit Ball
2 years 2 months ago

Not to get off topic but something tells me John Rocker is preparing to attend the funeral of the leader of the Westboro Baptist church.

Eric
Guest
Eric
2 years 2 months ago

this isn’t about smarts, I just want someone to answer the mailbag they provide, and they don’t.

After a month goes by you expect an answer.

If they aren’t going to answer it, then why offer it as a feature on the site?

Spit Ball
Guest
Spit Ball
2 years 2 months ago

Oh Eric….you poor thing.

Bearman
Guest
Bearman
2 years 2 months ago

Two whole emails? Damn ur persistent. 90% of the emails they get a day are probably from you.

Baltar
Guest
Baltar
2 years 2 months ago

That’s what the daily chats are for.

rmgxt9
Member
rmgxt9
2 years 2 months ago

email bag, he says.

tz
Guest
tz
2 years 2 months ago

I can only imagine what Pena’s batting average would be without those bunt hits.

And that, for me, is enough evidence that bunting success doesn’t force opponents’ hands to stop the shift. The GIFs of TB were just icing on the cake.

Spit Ball
Guest
Spit Ball
2 years 2 months ago

Stop the shift; no. Useful weapon; perhaps. Although I’d like to know what happened in the Plate Appearances that Pena attempted a bunt during but the bunt went foul. Small sample size but it would be neat to compare wOBA in those situations compared to situations where he did not try to bunt against a shift.

Muck Bartinez
Guest
Muck Bartinez
2 years 2 months ago

I seem to remember Pena attempting a lot of bunts against the shift, but many when there were two strikes against him. I’m remembering him swinging away or whatever for the first part of the at bat, but then after falling behind in the count (0-2, 1-2) attempting to bunt. I’m wondering if the Rays staff gave him some numbers like an OBP/wOBA with 2 strikes against a shift versus the odds of laying down a successful bunt with one try.
It makes for an interesting decision with two strikes, because obviously you only get one bunt try at that point, but we also know what happens to hitters numbers when they get to 0-2 or 1-2. Perhaps this avenue of research will give us an idea of whether it’s best to bunt against the shift in all counts, some counts, or none.

Spit Ball
Guest
Spit Ball
2 years 2 months ago

I find that interesting for another reason. If he’s bunting at 0-2 or 1-2 a bunch of the time he’s also trying to bunt breaking stuff which is not ideal.

n0exit
Guest
n0exit
2 years 2 months ago

I think it’s wrong to say bunting is hard by looking at these numbers. That’s just one possible interpretation. Another is that the guys in your sample aren’t good bunters. 65 attempts in 5 seasons doesn’t make you the bunting king against the shift (although you could argue that in the valley of the blind the man with one eye is king). Logic tells me if light hitters can become good bunters heavy hitters should be able to do the same. As for the effect on the shift, it drags one Felder over to third leaving a large hole up the middle. I would expect this to affect which pitches are thrown. If you can make the pitcher do something he doesn’t want to that has to be worth some spring training bunts.

bob
Guest
bob
2 years 2 months ago

Conditions of the field might have an effect. How steeply is the infield sloped to the foul lines, how long is the grass, how wet is the field? Would there need to a park factor for bunts?

I’ve seen bunt hits where the third baseman was not shifted but instead of simply letting the ball roll foul he unwisely fields it and throws late to first.

New stat: Batting Average on Bunts in Play?

J6takish
Guest
J6takish
2 years 2 months ago

Carlos Pena is the king of bunting agsinst the shift but also is the owner of a .233 batting average. I’m not sure where I’m going with this but I’m left wondering if a guy with his power was better off just going for it because bunting didn’t seem to help

MGL
Guest
MGL
2 years 2 months ago

This time, I don’t understand your numbers:

“Of those attempts, 38% were bunted fair, and 25% of the bunts resulted in the batter reaching base, either on a hit or an error. In other words, one of four attempted bunts put the batter on, but two of three bunts in play worked out, which isn’t a shock.”

So 38% were bunted fair and the other 62% were strikes I assume. The 25% number, though is confusing. 25% of what? Fair bunts? So the other 75% are outs? Or is the 25% of the attempts? So, we have 25% safe on a hit or error, 13% out, and 62% a strike? I assume that this is the case.

Well, that is quite a good result. Say you attempt a bunt on the first pitch. 25% of the time, you get a hit or an error, which is worth, say .45 runs. 13% of the time, you bunt into an out, which is worth -.25 runs. 62% of the time, you have an extra strike. What is the value of a PA with an extra strike? I say “extra” and not necessarily 0-1, because presumably you are only bunting if it is a good pitch, likely a strike, and you would pull back on many of the pitches out of the zone.

So, Jeff, you really need to tell us what happens when the ball is bunted foul. We need to know how the PA ends. Really all we need to know is how the PA ends when a batter attempts a bunt into the shift on the first pitch. Technically, we want to include the times that he was going to bunt, pulls back, and takes a ball or called strike, otherwise we are biased by PA that tend to start with a strike.

So, again, you numbers are not very informative and are in fact misleading. I am sure that I am not the only one that thought perhaps that the 25% means that when you bunt the ball in play, you are safe only 25% of the time.

So anyway, with an extra strike, I would guess that the PA is worth around -.025 runs or so. Just a wild, but educated guess.

So, we have .25 * .45 – .25 * .13 – .025 * .62, or:

.0645 for the value of the PA. That is equivalent to the value of a PA for a slugger like Fielder or Cabrera.

Swfcdan
Guest
Swfcdan
2 years 2 months ago

Good post, I thought that too.

Jonny Dollar
Guest
2 years 2 months ago

Enjoyed both of your bunting articles and think it’s great analysis that I’ve never personally seen before Jeff. Thanks for the work.

dsangdhw
Member
dsangdhw
2 years 2 months ago

Hello! everybody, give you recommend a good shopping place.

Carlos Pena
Guest
Carlos Pena
2 years 2 months ago
Baltar
Guest
Baltar
2 years 2 months ago

I still like the idea of taking a half-swing toward the huge hole.

roadrider
Guest
roadrider
2 years 2 months ago

When teams shift they’re conceding the bunt hit. Power hitters are doing the opposition a favor if they comply. That’s why the shift doesn’t go away even for guys who bunt a lot against it. Yeah its a good idea to drop one down (if you can) once in a while but I’d rather have my sluggers swing away.

Bryz
Guest
2 years 2 months ago

Let’s use Chris Davis as an example just for fun.

Last year with no one on base (let’s just assume this is the only scenario he faced a shift), he had 50 extra base hits in 339 AB, or about 15%. Jeff’s research showed that lefties that faced a shift and attempted a bunt laid it down in fair territory 38% of the time, and when it was successfully bunted fair, they succeeded in reaching base nearly 66% of the time.

I don’t know if the bonus of extra bases is worth the decrease in probability, especially when Davis is the extreme example from 2013.

Neil S
Guest
Neil S
2 years 2 months ago

“they’re conceding the bunt hit. Power hitters are doing the opposition a favor if they comply.”

The opposition might think that, but they’d be wrong. Giving a free base to a hitter is nearly always the wrong strategy. With 0 or 1 out, we’re talking about an increase in run expectancy between .2 and .4, which is pretty significant.

The only reason to concede the hit is if the following hitter is awful. Or, I suppose, if the guy you’re shifting against doesn’t bunt or doesn’t bunt well. In which case, you’re not actually conceding anything.

Schuxu
Guest
Schuxu
2 years 2 months ago

Following this logic, as the pitching team you should always just intentional walk the power hitters.

channelclemente
Guest
channelclemente
2 years 2 months ago

You’d think that with successful bunts, and fewer shifts likely, you’d see a hitter’s BABIP creep up.

Nivra
Guest
Nivra
2 years 2 months ago

It feels like some math is probably the next step in this investigation. We know the LWTS cost of an extra strike for various counts. We also know league average successful bunt rates, as well as bunt rates of various sluggers.

So, let’s do the math. What’s the break-even point of attempting a bunt on an 0-0 count? 40% fair? 30% fair?
What about with 2 strikes? 50% fair? 70% fair?

It obviously changes depending on the skill of the batter, but maybe start with a league average batter, then extend the analysis for wOBA’s above the league average?

Mariners fan
Guest
Mariners fan
2 years 2 months ago

Next somebody should check for pitch selection after shown/foul bunts. Do they affect the pitch sequence?

Swfcdan
Guest
Swfcdan
2 years 2 months ago

Biggest reason along with the increased strike zone size, for the decline in offense over recent years. Its happening to more and more batters, interesting that it’s so many lefties compared to righties.

The shifts have caused a big decline in batting average as the average batting average now is barely over .250, which was awful 5-10 years ago.

The teams have adjusted by laying on shifts on a regular basis, it’s now time for all sluggers who are being shifted to adjust too and learn bunting. It’s the only way these damn offense sapping shifts are going to go away. And even then I think offense will never get back to the level it was, roids aside…

It’s a pitching ERA now.

BernieWilliamsFan
Member
BernieWilliamsFan
2 years 2 months ago

One thing that might be interesting to look at is hitter performance during a defensive shift when bunting. That might be a good way to judge success during a bunt. If the hitters did better when not bunting, it might be worthwhile to just let them swing away. Just a thought.

wpDiscuz